Adaptive optical focusing through perturbed scattering media with a dynamic mutation algorithm

Abstract

Optical focusing and imaging through or inside scattering media, such multimode fiber and biological tissues, has significant impact in biomedicine yet considered challenging due to strong scattering nature of light. In the past decade, promising progress has been made in the field, largely benefiting from the invention of iterative optical wavefront shaping, with which deep-tissue high-resolution optical focusing and hence imaging becomes possible. Most of reported iterative algorithms can overcome small perturbations on the noise level but fail to effectively adapt beyond the noise level, e.g. sudden strong perturbations. Re-optimizations are usually needed for significant decorrelation to the medium since these algorithms heavily rely on the optimization performance in the previous iterations. Such ineffectiveness is probably due to the absence of a metric that can gauge the deviation of the instant wavefront from the optimum compensation based on the concurrently measured optical focusing. In this study, a square rule of binary-amplitude modulation, directly relating the measured focusing performance with the error in the optimized wavefront, is theoretically proved and experimentally validated. With this simple rule, it is feasible to quantify how many pixels on the spatial light modulator incorrectly modulate the wavefront for the instant status of the medium or the whole system. As an example of application, we propose a novel algorithm, dynamic mutation algorithm (DMA), which has high adaptability against perturbations by probing how far the optimization has gone toward the theoretically optimal performance. The diminished focus of scattered light can be effectively recovered when perturbations to the medium cause significant drop of the focusing performance, which no existing algorithms can achieve due to their inherent strong dependence on previous optimizations. With further improvement, the square rule and the new algorithm may boost or inspire many applications, such as high-resolution optical imaging and stimulation, in instable or dynamic scattering environments.

Anti-scattering light focusing by fast wavefront shaping based on multi-pixel encoded digital-micromirror device

“…Encoding multiple neighboring pixels in the DMD enables a quasi-continuous light amplitude modulation, circumventing the obstacles created by the conventional binary amplitude modulation, which unambiguously improves the competitiveness of DMD for WS. Moreover, SNES as a global search algorithm has a faster speed to figure out the optimal DMD modulation than the other conventional methods like the "half-blind" GA method [19][20][21][22][23][24][25] . In our experiments, we compared the mpDMD-SNES WS method with the state-of-the-art bDMD-GA method.…”

“…To achieve high-enhancement light focusing within highly dynamic scattering media, high-speed iterative optimization and a large number of modulation modes are essential for the iterative WS method. On the one hand, a variety of iterative optimization algorithms have been proposed, including the continuous sequential optimization 20,21 , partitioning algorithm 13 , genetic algorithm (GA) 19,22 , particle swarm optimization 23 , Hadamard algorithm 24 , simulated annealing algorithm 25 , etc. Compared with the conventional transmission matrix methods that require an enumerate search, these methods [19][20][21][22][23][24][25] significantly accelerate the process of optimizing the incident wavefront, but the speed of optimization is still not high enough to deal with the high dynamic media.…”

“…On the one hand, a variety of iterative optimization algorithms have been proposed, including the continuous sequential optimization 20,21 , partitioning algorithm 13 , genetic algorithm (GA) 19,22 , particle swarm optimization 23 , Hadamard algorithm 24 , simulated annealing algorithm 25 , etc. Compared with the conventional transmission matrix methods that require an enumerate search, these methods [19][20][21][22][23][24][25] significantly accelerate the process of optimizing the incident wavefront, but the speed of optimization is still not high enough to deal with the high dynamic media. In addition, these methods involve stochastic variables in the optimization and are sensitive to the initial condition and local minima, which constrains the enhancement of the focus.…”

Approaching the optimum focusing of diffused light through scattering media with iterative wavefront shaping by a probability-based algorithm

“…However, due to the inhomogeneities of refractive index in biological tissue, photons are multiply scattered in tissue sample and the amount of ballistic photons decays exponentially with increasing propagation depth, limiting high-resolution optical imaging to a depth of ~1 mm beneath the skin or tissue surface 1 . Thanks to the emergence of optical wavefront shaping techniques, scattering-induced wavefront distortions nowadays can be compensated via various approaches, such as optical phase conjugation [2][3][4][5][6][7][8] , transmission matrix method 9-12 , iterative optimization [13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28][29] , and artificialintelligence-assisted methods [30][31][32][33] , allowing optical focusing or imaging inside/through scattering media. Iterative optimization approaches are widely adopted because they are straightforward and less technically demanding.…”

Deep learning-empowered wavefront shaping in scattering media

“…The idea of hybridizing complementary algorithms can be extended to nonstatic scenarios. For example, the hybrid of the RFOTNet/TFOTNet and the dynamic mutation algorithm [133] will improve wavefront shaping efficiency through/within nonstatic media. The fine-tuning of RFOTNet demonstrated in Chapter 4 is proposed to cope with environmental perturbations, and it is a relatively simpler model compared with models used in randomly changing media.…”